Proposal for a common multimedia ontology framework: Information management for music analysis systems Samer A. Abdallah, Yves Raimond, Mark Sandler Centre for Digital Music, Queen Mary, University of London {samer.abdallah,yves.raimond,mark.sandler}@elec.qmul.ac.uk 1 Introduction In this proposal, we discuss requirements for and potential applications of a music information management system that represents in a richly structured way not only the media objects themselves and associated metadata, but also the computational systems by which new features and representations of given media objects can be derived. In the first instance, this is intended to support the activities of researchers, who may be developing new algorithms for analysis audio or symbolic represen- tations of music, or may wish to apply methodically a battery of such algorithms to a collection or multiple sub-collections of music. For example, we may wish to examine the performance of a number key finding algorithms on a varied collection, grouping the pieces of music along multiple dimentions by, say, in- strumentation, genre, and date of composition. The knowledge representation should support the definition of this experiment in a succinct way, selecting the pieces according to given criteria, applying each algorithm, perhaps multiple times in order to explore the algorithms’ parameter spaces, adding the results to the knowledge base, evaluating the performance by comparing the estimated keys with the annotated keys, and aggregating the performance measures by instrumentation, genre and date of composition. The outputs of each algorithm should be added to the knowledge base in such a way that each piece of data generated is unambigously associated with the function that created it and all the parameters that were used, so that the resulting knowledge base is fully self- describing. Finally, a statistical analysis could be performed to judge whether or not a particular algorithm has successfully captured the concept of ‘key’, and if so, to add this to the ontology of the system so that the algorithm gains a semantic value; subsequent queries involving the concept of ‘key’ would then be able to invoke that algorithm even if no key annotations are present in the knowledge base. One of the key ideas here is to blur the disctinction between a ‘relation’ or ‘property’ in a database or ontology and a ‘function’ provided by a software library—in many respects, a function is like a relation whose tuples are not stored explicitly, but are computed on demand when enough attributes (i.e. the inputs to the function) have been supplied. By making a complete record of